British Journal of Nutrition (2007), 97, 247–254
q The Authors 2007
DOI: 10.1017/S0007114507328614
Mononuclear cell metallothionein mRNA levels in human
subjects with poor zinc nutrition
Chong-Suk Kwon1, Aggeliki M. Kountouri2, Claus Mayer2, Margaret-Jane Gordon2,
In-Sook Kwun1 and John H. Beattie2*
1
2
Department of Food Science and Nutrition, Andong National University, Kyungpook, 760-749, South Korea
Division of Cellular Integrity, Rowett Research Institute, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, UK
(Received 24 May 2006 – Revised 17 August 2006 – Accepted 21 August 2006)
Human zinc deficiency is thought to be prevalent worldwide, particularly in populations with diets low in zinc and animal protein and high in
inhibitors of zinc absorption, such as phytic acid. Confirmation of zinc deficiency is, however, difficult in the absence of a reliable and sensitive
marker of zinc status. Under controlled conditions, T-lymphocyte metallothionein-2A (MT-2A) mRNA levels change in relation to zinc status and
the objective of the present study was to investigate whether these transcript levels could be related to dietary zinc intake, plasma zinc or other
biochemical parameters influenced by, or influencing, zinc metabolism in human subjects likely to be zinc deficient. Rural Koreans (n 110, age
50 –80 years) with a range of zinc and phytic acid dietary intake were recruited for the study and blood samples were analysed for plasma zinc,
HDL, LDL, a-tocopherol and thiobarbituric acid reactive substances, mononuclear cell (MNC) MT-2A mRNA, serum protein and albumin, and
blood haematocrit, Hb and glucose. Multiple correlation and principal component analysis showed a significant negative correlation between
plasma zinc and MNC MT-2A mRNA levels. Female subjects had higher MT-2A transcript levels than males and MT-2A mRNA levels
tended to increase with age. There was no significant association between dietary zinc intake or any index of zinc intake relating to dietary inhibitors of zinc absorption. It is concluded that MNC MT-2A mRNA levels cannot be used to predict poor zinc nutrition.
Mononuclear cells: Metallothionein: Zinc nutrition: Zinc status
The WHO estimates that over two billion people worldwide may
be zinc deficient (World Health Organization, 2002). This
assessment is based on food availability data not on zinc
status, which is difficult to assess because of the lack of reliable
status markers (Hambidge, 2003). Several diagnostic markers of
zinc status have been evaluated, including levels of plasma zinc,
plasma alkaline phosphatase, plasma thymulin and hair/nail zinc
(McKenzie, 1979; Wood, 2000). All of these markers have
proven unsatisfactory for various reasons, but the greatest disadvantages seem to be that they are not specific for zinc deficiency
and they have limited sensitivity. Metallothionein (MT) was
proposed as a marker of zinc status many years ago (Grider
et al. 1990; King, 1990) due to the sensitivity and magnitude
of its induction by zinc. Zinc-deficient rats had substantially
lower plasma levels of MT (Sato et al. 1984) and MT levels in
human erythrocytes decreased in experimental zinc deficiency
(Grider et al. 1990; Thomas et al. 1992). The problem with
measuring plasma MT was the lack of available immunoassays
with sufficient sensitivity and specificity. Analysis of erythrocyte MT was confounded by the observation that most of the
MT was concentrated in the reticulocyte fraction, and that
changes in erythropoesis could therefore influence total erythrocyte MT levels (Robertson et al. 1989). Mononuclear cell
(MNC) MT mRNA levels measured by competitive RT –PCR
methods have been shown to be positively correlated with zinc
status (Allan et al. 2000; Cao & Cousins, 2000), and the development of real-time RT–PCR has improved the ease with
which this target can be quantified.
There are several dietary factors that can affect zinc absorption, among the most notable being phytate and calcium,
which decrease absorption (Krebs, 2000; Lonnerdal, 2000).
Molar ratios of . 20 for the phytate:Zn content and . 200
for the phytate £ Ca:Zn content of diets have been recognized
as likely to induce zinc deficiency (Ruz et al. 1991), and the
diets of some population groups throughout the world exceed
these threshold ratios. Indeed, mean values of 40 and 400 have
been recorded for phytate:Zn and phytate £ Ca:Zn ratios,
respectively, in the diets of rural Koreans (Lee et al. 2004).
Plasma alkaline phosphatase levels in 176 volunteers at risk of
zinc deficiency according to their zinc, phytate and calcium
intakes, were found to be considerably lower than the reference
range, although plasma zinc was not lower than the respective
reference range. Although the utility of quantifying T-lymphocyte MT-2A mRNA for the assessment of zinc status had been
demonstrated in a controlled intervention study (Allan et al.
2000), its application to the assessment of zinc status in population groups has not been evaluated. We proposed therefore
to measure MT-2A mRNA levels in mononuclear leucocytes
from rural Korean subjects who were likely to show a range of
zinc intakes and status. Taking advantage of samples from an
Abbreviations: MNC, mononuclear cell; MT, metallothionein.
* Corresponding author: Dr John H. Beattie, fax þ44 (0)1224 716629, email [email protected]
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 17 Jun 2017 at 14:30:59, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/S0007114507328614
248
C.-S. Kwon et al.
on-going large study relating nutritional status to health in different Korean localities, we identified a sub-group of individuals
suitable for the present study. Our major objectives were twofold: (1) to investigate the range and variation of MT gene
expression in MNC from subjects in a population and to relate
the MT levels to variable factors such as age and sex, and (2)
to investigate whether dietary zinc intake and plasma zinc is
related to MNC MT expression.
Methods and materials
Volunteer recruitment
Subjects were recruited as part of the Public Medical Center
Health Promotion Program Project, from fifteen rural localities
in the Andong area of South Korea. Recruitment was achieved
by sending an information postcard to the residents of each
area and those wishing to join the project were accepted, subject to meeting the inclusion criteria. The final number of volunteers recruited to the study was around 1700, and they were
assumed to be representative of the whole population. Subjects
were aged 50 –80 years with a BMI within the normal range
(18–26). A sub-sample of 110 volunteers from seven localities
were selected for the present study on the basis of their zinc
intake. This sub-population was representative of the range
of zinc intakes found in the larger study group and contained
both male and female volunteers. Exclusion criteria included
any clinical signs of chronic disease or infections and any
long-term medication or supplementation. Ethical permission
for the present study was granted by the Andong Public Medical Center Ethical Committee.
Body weight ranged from 58 to 66 kg for men and 51 to
59 kg for women. Height ranged from 162 to 167 cm and
from 146 to 155 cm for men and women, respectively, and
BMI from 21·94 to 24·40 for males and from 23·90 to 25·61
for females. BMI was calculated by dividing body weight in
kg by (height)2 in m. The volunteer age and sex for each
locality are presented in Table 1.
Health status and dietary assessment
For screening medical status a general questionnaire was used
to obtain information on the medical history of subjects and
their family history, including the nature of any disease and
its duration. A routine physical examination was also made
by medical examiners of the Andong Public Medical Center
Health Promotion Program.
A FFQ was used for estimation of the intakes of zinc, calcium,
phytate and other nutrients. The FFQ, which contained thirtyeight food items, included the major food sources and major
zinc-containing food items which are commonly consumed by
Koreans during the four seasons of each year. FFQ data were correlated with 24 h recall information to validate the FFQ method.
Nutrient intakes were calculated using Computer Aided
Nutritional Analysis Program, version 2.0 (CAN Pro 2.0,
Korea Nutrition Society, Seoul, South Korea). Intake for
zinc and phytate, which are not included in the CAN Pro
2.0 program, were analysed using food composition tables,
databases, a cross-referenced index and various values from
the literature (Korean Food & Drug Association, 1996;
Korean Food and Drug Safety Section, 2002; Kwun &
Kwon, 2000). The procedure for calculation of zinc, calcium,
phytate and phytate:Zn molar ratio has been described previously (Kwun & Kwon, 2000). Nutrient intakes were compared with the Korean Recommended Dietary Allowance
(Korean Nutrition Society, 2000).
Blood sampling and processing
A fasted blood sample was collected from subjects in the early
morning, and plasma, erythrocytes and MNC were separated
by layering the blood sample on to Histopaque 1077 (Sigma
Aldrich, St Louis, MO, USA), and centrifuging at 400 g for
30 min. Plasma was removed and frozen at 2 808C while the
MNC at the Histopaque interface were removed and centrifuged
at 250 g for 10 min to obtain a cell pellet. The pellet was washed
with saline and the cells re-centrifuged and frozen at 2 808C in
1 ml RNALater (Ambion, Austin, TX, USA).
RNA isolation
The MNC samples were defrosted on ice and centrifuged at
3000 g for 15 min at 48C (MR 1822 Jouan centrifuge; ThermoElectron, Basingstoke, UK) in order to pellet them. The RNALater supernatant was discarded and RNA was extracted using
TRIzol according to the instructions of the manufacturer (Invitrogen, Paisley, UK). The RNA pellet was briefly dried for
5–10 min and then dissolved in 10 ml RNase-free water. Each
sample was diluted £ 1000 and quantified by absorbance at
260 nm using an Eppendorf Biophotometer (Eppendorf UK
Ltd, Cambridge, UK). The samples were stored at 2808C.
cDNA synthesis by reverse transcription of RNA
The RNA samples were diluted to a final concentration of
1 mg/ml. A mastermix containing 2 ml 10 £ concentrated RT
buffer, 4·4 ml 25 mM -MgCl, 4 ml 10 mM -DNTP (with dTTP)
and 1 ml 50 mM -random hexamers was prepared by mixing,
and 0·4 ml RNase inhibitor (20 units/ml) and 1 ml Multiscribe
RT (50 units/ml) (Applied Biosystems, Warrington, UK)
were subsequently added, also with mixing. To obtain a
20 ml reaction mix, 5 ml RNA sample and 2·2 ml water were
added to 12·8 ml of the mastermix. The reaction mixes were
put into a thermal cycler (Hybaid Omn-E, Ashford, UK) and
were heated at 308C for 10 min, then at 488C for 30 min and
finally at 958C for 5 min. The samples were frozen at
2 808C until required for the human MT-2A PCR.
Real-time PCR for human metallothionein-2A
After reverse transcription, MT-2A cDNA was amplified by realtime PCR using the TaqMan system (Applied Biosystems). PCR
reaction mixes were prepared in a PCR workstation and reagents
were dispensed using positive-displacement pipettes with autoclaved, UV-irradiated tubes and tips. cDNA samples were
defrosted in ice and a premix of 50 ml/tube of TaqMan Universal
Master Mix (Applied Biosystems), 35 ml RNA-free water and
6 ml and 3 ml/tube MT-2A for the forward –reverse primers
and probe, respectively, were prepared. The purified primers
were obtained from MWG-Biotech AG (London, UK) and the
probe from Applied Biosystems. The sequences of the probes
and primers for MT-2A were:
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 17 Jun 2017 at 14:30:59, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/S0007114507328614
(Mean values with their standard errors)
Age
Rural locality
1
1
2
2
3
3
4
4
5
5
6
6
7
7
All
All
All
Energy intake
(kJ/d)
Protein intake
(g/d)
Protein intake
(% animal
protein)
Zn intake
(mg/d)
Zn:LBM ratio
Phytate:Zn
ratio
Phytate £
Ca:Zn ratio
Plasma Zn
(mm)
MT mRNA
Sex
n
Mean
SE
Mean
SE
Mean
SE
Mean
SE
Mean
SE
Mean
SE
Mean
SE
Mean
SE
Mean
SE
Mean
SE
M
F
M
F
M
F
M
F
M
F
M
F
M
F
M
F
All
9
9
7
9
5
4
12
10
4
8
8
15
6
4
51
59
110
60·2
59·1
67·7
65·8
59·4
56·8
61·4
62·7
59·3
63·9
63·9
60·7
63·7
65·8
62·4
62·1
62·2
2·2
1·7
2·4
2·0
2·8
3·1
1·8
2·2
2·4
2·5
3·3
1·8
2·4
2·5
1·0
0·9
0·6
6879
6649
6875
5673
6033
7336
6628
6050
6100
6029
7055
6214
6155
6084
6649
6184
6364
2638
883
657
1034
636
1833
997
1445
406
1097
1189
1859
896
988
1428
1369
1403
60·1
53·0
54·4
41·1
46·6
69·0
57·2
51·0
48·4
47·0
55·7
56·4
46·8
48·6
54·8
51·3
52·6
23·0
10·3
7·1
10·0
11·0
19·6
15·0
15·7
6·3
13·2
11·6
22·5
11·7
12·4
14·7
16·5
15·9
30·5
20·1
24·6
14·4
23·3
17·3
27·2
23·0
24·1
19·2
20·3
26·5
18·8
17·6
25·4
21·2
22·8
9·9
6·7
9·7
8·6
9·2
20·9
12·3
7·9
9·9
11·0
9·4
15·0
11·8
9·1
10·4
11·3
11·1
8·4
6·6
6·2
6·0
7·6
8·7
7·6
6·0
6·4
7·0
5·6
6·7
6·6
5·2
7·0
6·5
6·8
1·4
0·6
0·4
0·7
0·9
1·7
0·9
0·6
1·9
1·5
0·5
1·0
0·5
0·4
0·4
0·4
0·3
2·78
2·68
2·10
2·60
2·48
3·23
2·39
2·47
2·19
2·85
1·76
2·84
2·14
2·53
2·29
2·72*
2·53
0·46
0·17
0·2
0·33
0·27
0·54
0·28
0·25
0·49
0·65
0·15
0·41
0·19
0·26
0·13
0·15
0·10
21·4
22·2
27·3
19·3
21·9
13·2
24·1
28·1
26·0
19·4
29·4
26·1
23·2
26·5
24·7
23·1
23·8
2·7
3·7
5·2
1·6
3·8
1·8
1·9
3·1
2·5
2·5
1·6
2·5
2·1
2·1
1·1
1·2
0·8
256·1
198·2
412·1
143·2
235·9
112·8
256·7
259·0
192·6
143·8
244·7
260·2
371·1
226·1
282·4
204·6
240·7
46·5
57·2
142·3
33·2
88·5
11·9
45·9
52·3
33·5
40·4
53·1
36·6
60·5
50·0
27·7
18·3
16·5
14·4
11·7
14·9
15·2
14·6
10·8
11·1
9·6
11·9
11·4
10·6
10·5
12·6
13·6
12·7
11·6
12·1
1·1
0·9
1·5
0·7
0·7
0·3
0·9
0·8
0·7
1·0
0·9
0·8
0·8
1·8
0·4
0·4
0·3
0·47
0·49
0·27
0·47
1·05
0·59
0·71
1·9
1·07
1·47
1·41
2·41
2·13
ND
0·85
1·33*
1·10
0·21
0·31
0·08
0·16
0·45
0·31
0·19
0·50
0·42
0·66
0·33
0·45
1·48
0·13
0·21
0·13
Metallothionein mRNA levels and poor zinc nutrition
F, female; LBM, lean body mass; M, male; MT, metallothionein; ND, not determined.
* Statistical significance (P # 0·05) when comparing males and females for all localities.
† For details of procedures, see pages 248 and 250.
249
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 17 Jun 2017 at 14:30:59, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/S0007114507328614
Table 1. Age, zinc intake, plasma zinc and mononuclear cell MT-2A mRNA (MT mRNA) in subjects according to rural locality (see p. 248) and gender. Zinc intake is also expressed as a proportion of
lean body mass (Zn/LBM), as a molar ratio with dietary phytic acid (Phy:Zn) and as a molar ratio index with dietary calcium and phytic acid (Phy x Ca:Zn)†
250
C.-S. Kwon et al.
Forward primer: 50 -GCA CCT CCT GCA AGA AAA GC-30
Reverse primer: 50 -TGG AAG TCG CGT TCT TTA CAT
CT-30
Probe: 50 -FAM-CTC CTG CTG CCC TGT GGG CTG TTAMRA-30
A single PCR product of 154 bp was consistently obtained,
with no evidence of genomic DNA amplification, which
would yield a product of 359 bp. 18S ribosomal RNA
(TaqMan Ribosomal RNA Control Reagents VIC Probe Protocol; Applied Biosystems) was used as a reference.
In each plate one non-template control and seven standards
were run with the samples. Since the objective was to measure
relative rather than absolute concentrations, standards were
made by diluting a selected sample, according to the protocols
supplied by Applied Biosystems. The concentration of
reverse-transcribed RNA in the top standard was 250 ng/ml,
and six serial dilutions were made to give the following
additional standard concentrations: 62·5, 15·62, 3·9, 0·97, 0·24
and 0·04 ng/ml; and for 18S: 10, 2·5, 0·625, 0·156, 0·039, 0·010
and 0·003 ng/ml. The samples were diluted to 50 ng/ml. To
each microcentrifuge tube, 51·8 ml premix and 2·2 ml of water,
standards or sample cDNA were added. The tubes were vortexed
and then centrifuged for 1 min at high speed using a Sigma 1–13
centrifuge. Then 2 £ 25 ml of control, standards and samples
were dispensed into duplicate wells of a PCR ninety-six-well
plate. To remove any bubbles in the wells, the plate was sealed
and centrifuged for 1 min in a Sigma 3–15 centrifuge. The
plate was then analysed for target transcript levels by real-time
PCR using an Applied Biosystems PRISM 7700 instrument.
Standard operating conditions of 508C for 2 min (for Uracil
N-glycosylase activity), 958C for 10 min (for hot-start taq polymerase activation) and forty cycles of 958C for 15 s (denaturation) and 608C for 1 min (annealing and extension) were used.
(Fig. 4). Correlations were calculated using the Pearson correlation coefficient. For the multiple regression analysis, a stepwise regression method was used to find the appropriate
model. The ‘Stepwise’ procedure of the computer program
S-Plus 2000 (1999, MathSoft Inc., Cambridge, MA, USA)
which is based on the Efroyson-algorithm (Miller, 1979),
was used with default settings.
Results
Zinc intake
The mean zinc intake for males and females in all rural
localities studied was estimated to be 7·0 and 6·5 mg/d,
respectively (Table 1). However, when corrected for lean
body mass, zinc intake was significantly higher in females
than males (Table 1). The range of zinc intake was very
wide, being 1·7 –20·6 mg/d, although over 75 % of all subjects
had an intake of between 4 and 8 mg/d (Fig. 1). The phytate:Zn ratio was in excess of 20, and the phytate £ Ca:Zn
ratio was in excess of 200 for most localities (Table 1).
Plasma zinc and mononuclear cell metallothionein-2A mRNA
As with zinc intake, plasma zinc showed a wide variation
ranging from 6·5 to 20·4 mM (Fig. 2). The overall mean for
110 subjects was 12·1 mM , with males having slightly higher
levels (12·7 mM ) than females (11·6 mM ). The levels of MNC
MT-2A mRNA were 1·5-fold higher in females than in
males (P¼0·05). There was considerable locality-related variation in MT-2A mRNA levels (Table 1) and they tended to
increase with age for both male and female subjects, although
this trend was not statistically significant (Fig. 3(A)). There
was a significant negative association between MT-2A
mRNA and plasma zinc (Fig. 3(B)).
Additional blood analyses
Statistical analysis
A correlation-based principal component analysis was performed on the study variables and the relationship between
the variables was then visualized in a corresponding bi-plot
Principal component analysis
Number of subjects
Figure 4 shows the bi-plot of the principal component analysis
of all the study variables and highlights and summarizes all
interactions in one figure, including those mentioned earlier.
Notable positive associations were between plasma zinc and
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
1·
0
2· – 1
0 ·9
3· – 2
0 ·9
4· – 3
0 ·9
5· – 4
0 ·9
6· – 5
0 ·9
7· – 6
0 ·9
8· – 7
0 ·9
9· – 8
10 0 – ·9
9
·
11 0 – ·9
·0 10
12 – ·9
· 1
13 0 – 1·9
· 1
14 0 – 2·9
· 1
15 0 – 3·9
· 1
16 0 – 4·9
· 1
17 0 – 5·9
· 1
18 0 – 6·9
· 1
19 0 – 7·9
· 1
20 0 – 8·9
·0 19
– ·9
20
·9
Blood haematocrit was measured by the microhaematocrit
method using heparinized capillary tubes and a dedicated centrifuge. Hb, plasma protein, albumin, glucose and HDL were
measured using commercial assay kits (Asan Pharmaceuticals,
Seoul, South Korea), and LDL was calculated using the Friedwald formula. Plasma thiobarbituric acid reactive substances
were determined by the method of Yagi (1984) and plasma
a-tocopherol was measured by the method of Bieri et al.
(1979) using reversed-phase HPLC (Shimadzu Corporation,
Kyoto, Japan). Plasma zinc was determined on duplicate
samples using atomic absorption spectrophotometry (Solaar
SP9, Unicam, UK) after 5-fold dilution with 0·1 M -HCl and centrifugation. The accuracy of zinc determination was verified by
frequent analysis of a standard reference material (Trace
Elements, Serum Level 1; Seronorm, Sero, Norway) which
was prepared by dissolving the freeze-dried powder in ultrapure
water, as instructed and diluting 5-fold with 0·1 M -HCl followed
by centrifugation. Plasma homocysteine levels were analysed by
an isotope dilution GC–MS technique (Calder et al. 1999).
Zn intake (mg/d)
Fig. 1. Zinc intake frequency distribution profile of volunteers (A, males; B,
females) from all localities. For details of procedures, see p. 248.
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 17 Jun 2017 at 14:30:59, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/S0007114507328614
Metallothionein mRNA levels and poor zinc nutrition
251
(A)
2·0
1·8
1·6
6
MT-2A mRNA
Number of subjects
8
4
2
1·4
1·2
1·0
0·8
0·6
0·4
0
6·
0
–
7· 6·9
0
–
8· 7·9
0
–
9· 8·9
0
10 – 9
·9
·0
11 – 1
·0 0·9
12 – 1
·0 1·9
13 – 1
·0 2·9
14 – 1
·0 3·9
15 – 1
·0 4·9
16 – 1
·0 5·9
17 – 1
·0 6·9
18 – 1
·0 7·9
19 – 1
·0 8·9
20 – 1
·0 9·9
–
20
·9
0·2
0·0
50–59
Plasma Zn (µM)
plasma glucose, and dietary zinc intake and plasma glucose.
Negative associations included dietary zinc intake and
phytate:Zn ratio. A highly significant positive correlation
(R 0·53, P,0·001) between dietary zinc intake and phytate
intake was observed (data not shown).
Multiple regression analysis
In the multiple regression analysis, the influence of all
measured variables on the variation in plasma zinc and
MNC MT mRNA levels was evaluated. The most influential
of these variables are shown in Table 2 in order of decreasing
statistical significance. In the case of plasma zinc, the most
influential variable was the serum albumin level, and in the
case of MNC MT-2A mRNA it was the plasma zinc level.
70+
Age (years)
(B)
MT-2A mRNA
Fig. 2. Plasma zinc frequency distribution profile of volunteers (A, males; B,
females) from all localities. For details of procedures, see p. 250.
60–69
2·2
2·0
1·8
1·6
1·4
1·2
1·0
0·8
0·6
0·4
0·2
0·0
<9
9–11
11–13
13–15
15–17
17+
Plasma zinc (µM)
Discussion
Fig. 3. Levels of mononuclear cell (MNC) metallothionein-2A (MT-2A) mRNA
related to volunteer age (A; B, males; A, females) and plasma zinc (B). For
details of procedures, see pp 248 and 250. Values are means with their standard errors depicted by vertical bars. Regression of plasma zinc with MNC
MT-2A mRNA gave a correlation coefficient (R) of 2 0·296 (P¼ 0·004) but
there was no significant correlation of MT-2A mRNA with age (P¼0·111).
Dietary zinc intake by rural Koreans is known to be below what
is considered adequate, and the risks of zinc deficiency are compounded by high dietary intake of phytic acid and a low intake of
animal protein (Kwun & Kwon, 2000). Phytate is a tenacious
chelator of zinc and reduces zinc absorption (Lo et al. 1981;
Krebs, 2000; Lonnerdal, 2000) and high molar ratios of phytate:Zn in human populations have been associated with zinc
deficiency (Ruz et al. 1991). In the present study, phytate:Zn
molar ratios were . 15 in all but one locality, indicating low
zinc absorption (Hotz et al. 2003). Dietary zinc intake (mean
intake of 7·0 and 6·5 mg Zn/d for males and females, respectively) was considerably lower than in metropolitan Korean
localities (about 11 mg Zn/d; Kwun & Kwon, 2000; Lee et al.
2004) and also in western countries such as the UK (mean
intake of 10·2 and 7·4 mg Zn/d for males and females, respectively) (Henderson et al. 2003). Phytic acid intake showed a
very significant positive correlation with zinc intake because
they both tend to be found in the same foods, particularly
grain, soyabean and its products, all of which are common constituents of Korean diets. However, principal component analysis indicated a negative correlation between phytate/zinc and
zinc intake, suggesting that the ratio of phytate to zinc was
higher at low zinc intake. This would indicate that individuals
most at risk of zinc deficiency through low consumption of
zinc may be at even higher risk due to poor zinc absorption efficiency. Calcium is also a dietary antagonist of zinc absorption
and a high index, calculated from multiplying molar levels of
calcium £ phytate and dividing by zinc, may induce zinc
deficiency (Ruz et al. 1991). The significance of calcium influences on zinc absorption, particularly in view of its binding affinity for phytic acid and the potential to compete with zinc
binding, remains controversial (Lonnerdal, 2000). Subjects in
most Korean localities had a phytate £ Ca:Zn index of . 200,
with some individuals having an index of 500–1000. Once
again, individuals with a lower zinc intake tended to consume
a higher proportion of Ca:Zn and phytate:Zn. On the basis of previous evaluations of dietary zinc intake and the impact of absorption inhibitors (Krebs, 2000; Lonnerdal, 2000), it seems likely
that most of the subjects recruited for the present study were
zinc deficient. Unfortunately, there are no reliable markers of
zinc status, but previous studies have demonstrated that rural
Koreans have significantly lower levels of plasma alkaline phosphatase, a zinc-dependent enzyme, than found in urban Koreans
(Lee et al. 2004).
Plasma zinc levels were not significantly related to dietary
zinc intake, whether expressed as mg Zn/d or as a ratio with
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 17 Jun 2017 at 14:30:59, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/S0007114507328614
252
C.-S. Kwon et al.
Haematocrit
Second component
0·2
0·1
Haemoglobin
Albumin
Plasma Zn
Phy x Ca/Zn
TBARS
Pro
Phy/Zn
0·0
HDL
α-Tocopherol
Glucose
LDL
Dietary Zn
Dietary Zn/LBM
Age
– 0·1
MT
– 0·2
– 0·3
– 0·2
– 0·1
0·0
0·1
0·2
0·3
First component
Fig. 4. Principal component analysis of study variables. The central point
from which all lines radiate represents a score of zero for both principal components. Lines radiating in a similar direction are made by variables that
show positive associations whereas those radiating in opposite directions are
made by variables that are negatively correlated. For example, mononuclear
cell metallothionein-2 mRNA (MT) is negatively correlated with plasma Zn
(Fig. 3(B)). Glucose refers to plasma glucose. LBM, lean body mass; Phy,
dietary phytate intake; Pro, plasma total protein; TBARS, plasma thiobarbituric acid reactive substances.
phytate and/or calcium, and were similar to those reported previously for Korean adults (Paik et al. 1999; Andree et al.
2004; Lee et al. 2004). Since it is highly likely that some of
the subjects in the present study were zinc deficient and
others were zinc adequate, the lack of responsiveness of
plasma zinc to varying zinc status is consistent with the current consensus of opinion (Wood, 2000; Hambidge, 2003).
The multiple regression analysis indicated that plasma zinc
was most influenced by serum albumin levels, followed by
blood Hb, blood glucose and HDL. The variables identified
by the analysis accounted for only 31 % of the variance in
the plasma zinc data, and so other unaccounted for influences
were also present. The positive association of plasma zinc
with serum albumin could relate to the role of this protein
in zinc transport, but considering that albumin is present in
considerable molar excess to the bound zinc, an explanation
based on simple binding capacity is not credible. The positive
association of plasma zinc with blood Hb was also quite
strong, and a positive association between Hb and haematocrit
was also apparent from the principal component analysis. This
suggests that erythrocyte counts may be reduced when plasma
zinc is low, and the observation that chronically zinc-deficient
mice become anaemic due to an attenuation of erythropoesis
(King et al. 2005) would seem to support this hypothesis.
Anaemia has previously been associated with zinc deficiency
but has been attributed to concomitant high dietary phytate
intake, which also inhibits iron absorption. The positive
relationship between blood glucose and plasma zinc is contrary to what might be expected from the known inverse
relationship between these variables (Sondergaard et al.
2006). Zinc stimulates insulin secretion (Huber & Gershoff,
1973) and receptor activation (Haase & Maret, 2003) which
facilitates glucose uptake by cells and elicits a decrease in
blood glucose levels. Acute zinc deficiency is reported to
reduce circulating levels of homocysteine in rats (Hong et al.
2000) but there was no evidence for any association between
plasma homocysteine and either dietary zinc intake, plasma
zinc or MNC MT mRNA levels (data not shown).
The strongest association of MNC MT-2A mRNA levels was
with plasma zinc levels, but it should be noted that the negative
correlation, which was highly significant for male and female
data collectively (P¼0·004), was much more significant in
women (P¼0·02) than in men (P¼ 0·22) (data not shown). The
inverse relationship could be explained in several different
ways. Since zinc stimulates MT gene expression, a higher
plasma level of zinc may be expected to promote an increase
in MT mRNA within cells directly exposed to blood plasma.
However, in the long term, the presence of adequate MT protein
levels may feed back to down-regulate gene expression. Alternatively, stress factors or the presence of inflammatory cytokines
may stimulate MT expression in tissues including MNC and
the liver, which would promote transfer of zinc from the
plasma into the tissues. This is a well-characterized response
of plasma zinc to stress (Shenkin, 1995) and is absent in MTdeficient mice (Rofe et al. 1996). Another possible explanation
is an influence of zinc on MNC composition. Lymphocytes,
monocytes and granulocytes express different levels of MT
(Kimura, 1991; Aydemir et al. 2006) and a zinc-related
change in the proportion of cell type may affect apparent MT
gene expression. A suppressive effect of zinc deficiency on lymphopoiesis and the reverse effect on myelopoiesis has been
Table 2. Factors which were most influential on the variation in plasma zinc and mononuclear cell MT-2A mRNA,
determined using stepwise multiple correlation. Explanatory variables are listed according to their significance and
R2 indicates how much of the response variation is explained by these factors (100 % when R2 ¼ 1)†
Response variable
2
Plasma zinc (R 0·308)
Metallothionein-2A mRNA (R 2 0·249)
Explanatory variable
Albumin
Hb
Glucose
HDL
a-Tocopherol
Intercept
Plasma Zn
LDL
Protein
Age
a-Tocopherol
Haematocrit
Intercept
Value
SE
t
P
1·6831
0·5241
0·0259
2 0·076
2 0·064
0·3478
2 0·101
2 0·008
0·4081
0·0353
0·0236
2 0·039
2 0·413
0·5606
0·1876
0·0103
0·031
0·0325
3·0737
0·0383
0·0037
0·1993
0·0178
0·0143
0·025
2·018
3·0023
2·7937
2·5078
2 2·4373
2 1·9565
0·1132
2 2·6257
2 2·236
2·0473
1·9814
1·6447
2 1·5754
2 0·2045
0·0034
0·0062
0·0137
0·0165
0·0531
0·9101
0·0103
0·028
0·0437
0·0508
0·1037
0·1189
0·8385
† For details of procedures, see p. 250.
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 17 Jun 2017 at 14:30:59, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/S0007114507328614
Metallothionein mRNA levels and poor zinc nutrition
reported (Fraker & King, 2004). However, given the wide
differences in MT expression that we observed and the predominance of lymphocytes in MNC fractions, it seems unlikely that a
change in the proportion of cell type could entirely explain the
relationship between MT mRNA and plasma zinc. However,
in the present study, it was impracticable to either separate the
different cell types or quantify their numbers, so we are unable
to discount this possibility.
A weak negative correlation (R 2 0·192) between MNC MT2A mRNA and plasma LDL was not statistically significant at
the 5 % level (P¼0·067). However, by multiple regression,
LDL was a significant contributory variable influencing the variance of MNC MT-2A mRNA levels (P¼0·028). This apparent
relationship requires further study for confirmation.
Of particular interest were the significantly higher levels of
MNC MT-2A mRNA in women compared to men and the
age-related increase in MT expression. The latter relationship
was not statistically significant but age was identified as a significant factor influencing MT-2A mRNA levels from the stepwise
multiple regression analysis and is worthy of further study. If
MT expression were only dependent on plasma zinc, then very
little age-related change in MT levels might be expected. However, blood levels of cytokines such as IL-6, which are also
known inducers of MT (Schroeder & Cousins, 1990), increase
with age (Krabbe et al. 2004) and it is likely that the age-related
increase in MT expression seen in the present study is a consequence of increased inflammatory stress and/or medication in
the elderly. As a consequence, plasma zinc levels might be
expected to decrease with age, but this was not the case in the
present study. There was no significant correlation between
age and plasma zinc (P¼0·464).
In conclusion, many of the rural Korean subjects recruited
for the present study were likely to be zinc deficient according
to their low dietary zinc intake and high intake of zinc absorption inhibitors. Plasma zinc was within the normal range for
adults and was unrelated to age or gender while MNC MT2A mRNA showed a weak age-related increase and also
higher levels in female subjects. A stronger inverse relationship between MT mRNA levels and plasma zinc might indicate that induction of MT in MNC by an unidentified factor,
such as inflammatory cytokines, could favour transfer of
zinc from the plasma to the tissues. MT mRNA levels did
not correlate with any measure of dietary zinc intake, and so
either the latter is a poor predictor of zinc status or MNC
expression of MT-2A cannot be used as a diagnostic indicator
of zinc status in free-living older human subjects.
Acknowledgements
J. H. B., A. M. K. and C. M. were funded by the Scottish
Executive Environment and Rural Affairs Department. C. S.
K. was supported by a grant from the Korean Science and
Engineering Foundation (grant number F01-2004-00010 277-0). I. S. K. was supported by the Korean Ministry of
Health and Welfare (grant number 03-PJ1-PG3-22 000-0038).
References
Allan AK, Hawksworth GM, Woodhouse LR, Sutherland B, King JC
& Beattie JH (2000) Lymphocyte metallothionein mRNA responds
253
to marginal zinc intake in human volunteers. Br J Nutr 84,
747–756.
Andree KB, Kim J, Kirschke CP, Gregg JP, Paik H, Joung H, Woodhouse L, King JC & Huang L (2004) Investigation of lymphocyte
gene expression for use as biomarkers for zinc status in humans.
J Nutr 134, 1716 –1723.
Aydemir TB, Blanchard RK & Cousins RJ (2006) Zinc supplementation of young men alters metallothionein, zinc transporter, and
cytokine gene expression in leukocyte populations. Proc Natl
Acad Sci USA 103, 1699 –1704.
Bieri JG, Tolliver TJ & Catignani GL (1979) Simultaneous determination of alpha-tocopherol and retinol in plasma or red cells by
high pressure liquid chromatography. Am J Clin Nutr 32,
2143 –2149.
Calder GM, Garden KE, Anderson SE & Lobley GE (1999) Quantitation of blood and plasma amino acids using isotope dilution electron impact gas chromatography/mass spectrometry with U-(13)C
amino acids as internal standards. Rapid Commun Mass Spectrom
13, 2080 –2083.
Cao J & Cousins RJ (2000) Metallothionein mRNA in monocytes and
peripheral blood mononuclear cells and in cells from dried blood
spots increases after zinc supplementation of men. J Nutr 130,
2180 –2187.
Fraker PJ & King LE (2004) Reprogramming of the immune system
during zinc deficiency. Annu Rev Nutr 24, 277–298.
Grider A, Bailey LB & Cousins RJ (1990) Erythrocyte metallothionein as an index of zinc status in humans. Proc Natl Acad Sci
USA 87, 1259 –1262.
Haase H & Maret W (2003) Intracellular zinc fluctuations modulate
protein tyrosine phosphatase activity in insulin/insulin-like growth
factor-1 signaling. Exp Cell Res 291, 289–298.
Hambidge M (2003) Biomarkers of trace mineral intake and status.
J Nutr 133, Suppl. 3, 948S – 955S.
Henderson L, Irving K, Gregory J, Bates CJ, Prentice A, Perks J,
Swan G & Farron M (2003) The National Diet & Nutrition
Survey: Adults Aged 19 to 64 Years; Vitamin and Mineral Intake
and Urinary Analytes. London: HMSO.
Hong KH, Keen CL, Mizuno Y, Johnston KE & Tamura T (2000)
Effects of dietary zinc deficiency on homocysteine and folate
metabolism in rats. J Nutr Biochem 11, 165– 169.
Hotz C, Lowe NM, Araya M & Brown KH (2003) Assessment of the
trace element status of individuals and populations: the example of
zinc and copper. J Nutr 133, 1563S–1568S.
Huber AM & Gershoff SN (1973) Effect of zinc deficiency in rats on
insulin release from the pancreas. J Nutr 103, 1739 – 1744.
Kimura M (1991) Metallothioneins of monocytes and lymphocytes.
Methods Enzymol 205, 291– 302.
King JC (1990) Assessment of zinc status. J Nutr 120, Suppl. 11,
1474 –1479.
King LE, Frentzel JW, Mann JJ & Fraker PJ (2005) Chronic zinc
deficiency in mice disrupted T cell lymphopoiesis and erythropoiesis while B cell lymphopoiesis and myelopoiesis were maintained. J Am Coll Nutr 24, 494– 502.
Korean Food and Drug Association (1996) Korean Food Composition
Table, 1996. Seoul: Korean Food and Drug Association.
Korean Food and Drug Safety Section (2002) Korean Food Composition Table, 2002. Seoul: South Korean Ministry of Health and
Welfare.
Korean Nutrition Society (2000) Korean Recommended Dietary
Allowance, 7th ed. Seoul: Korean Nutrition Society.
Krabbe KS, Pedersen M & Bruunsgaard H (2004) Inflammatory
mediators in the elderly. Exp Gerontol 39, 687– 699.
Krebs NF (2000) Overview of zinc absorption and excretion in the
human gastrointestinal tract. J Nutr 130, 1374S– 1377S.
Kwun IS & Kwon CS (2000) Dietary molar ratios of phytate: zinc and
millimolar ratios of phytate calcium: zinc in South Koreans. Biol
Trace Elem Res 75, 29– 41.
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 17 Jun 2017 at 14:30:59, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/S0007114507328614
254
C.-S. Kwon et al.
Lee SL, Kwak EH, Yoon JS, Kwon CS, Beattie JH & Kwun IS (2004)
Zinc nutritional status in Korean adults from rural, urban and
metropolitan areas. J Food Sci Nutr (Korea) 9, 174 – 182.
Lo GS, Settle SL, Steinke FH & Hopkins DT (1981) Effect of phytate:zinc molar ratio and isolated soybean protein on zinc bioavailability. J Nutr 111, 2223 – 2235.
Lonnerdal B (2000) Dietary factors influencing zinc absorption. J Nutr
130, 1378S –1383S.
McKenzie JM (1979) Content of zinc in serum, urine, hair, and toenails of New Zealand adults. Am J Clin Nutr 32, 570 – 579.
Miller A (1979) Subset Selection in Regression. London: Chapman
and Hall/CRC.
Paik HY, Joung H, Lee JY, Lee HK, King JC & Keen CL (1999)
Serum extracellular superoxide dismutase activity as an indicator
of zinc status in humans. Biol Trace Elem Res 69, 45– 57.
Robertson A, Morrison JN, Wood AM & Bremner I (1989) Effects of
iron deficiency on metallothionein-I concentrations in blood and
tissues of rats. J Nutr 119, 439 – 445.
Rofe AM, Philcox JC & Coyle P (1996) Trace metal, acute phase and
metabolic response to endotoxin in metallothionein-null mice. Biochem J 314, 793–797.
Ruz M, Cavan KR, Bettger WJ, Thompson L, Berry M & Gibson RS
(1991) Development of a dietary model for the study of mild zinc
deficiency in humans and evaluation of some biochemical and
functional indices of zinc status. Am J Clin Nutr 53, 1295 – 1303.
Sato M, Mehra RK & Bremner I (1984) Measurement of plasma
metallothionein-I in the assessment of the zinc status of zincdeficient and stressed rats. J Nutr 114, 1683 –1689.
Schroeder JJ & Cousins RJ (1990) Interleukin 6 regulates metallothionein gene expression and zinc metabolism in hepatocyte
monolayer cultures. Proc Natl Acad Sci USA 87, 3137– 3141.
Shenkin A (1995) Trace elements and inflammatory response: implications for nutritional support. Nutrition 11, 100– 105.
Sondergaard LG, Stoltenberg M, Doering P, Flyvbjerg A & Rungby J
(2006) Zinc ions in the endocrine and exocrine pancreas of zinc
deficient rats. Histol Histopathol 21, 619–625.
Thomas EA, Bailey LB, Kauwell GA, Lee DY & Cousins RJ (1992)
Erythrocyte metallothionein response to dietary zinc in humans.
J Nutr 122, 2408 –2414.
World Health Organization (2002) Quantifying selected major risks
to health. In The World Health Report 2002, pp. 47– 97. Geneva:
WHO.
Wood RJ (2000) Assessment of marginal zinc status in humans. J Nutr
130, 1350S– 1354S.
Yagi K (1984) Assay for blood plasma or serum. Methods Enzymol
105, 328–331.
Downloaded from https:/www.cambridge.org/core. IP address: 88.99.165.207, on 17 Jun 2017 at 14:30:59, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/S0007114507328614